Condom use predictors for low-income African American women.

The purpose was to investigate whether an HIV-prevention curriculum and a health maintenance curriculum produced different rates of change in reported consistent condom use and to explore what mediating variables predicted reported consistent condom use over time. A longitudinal crossover research design with extended posttest observations was used with a sample of 279 African American women. After pretesting, women received the designated curriculum and were posttested at completion, 3, 6, and 9 months. After the 9-month posttest, the curriculum was switched, and the above procedure was repeated. Using SAS macro GLIMMIX, the data revealed that the HIV curriculum yielded higher consistent condom use than did the health maintenance curriculum and that the mediating variables that predicted consistent condom use over time were self-efficacy for low-risk HIV behavior, HIV-related community behavior, and social norms. Enhancing consistent condom use over time may require the promotion and reinforcement of these mediating variables.